The non-extensivity parameter q in Tsallis statistics has been used to fit earthquake frequency-magnitude distributions for decades. It works well — q values between 1.4 and 1.8 reproduce the power-law tails observed in global seismicity data. But q has always been treated as a free parameter, adjusted to match each dataset. Sotolongo-Costa and collaborators show it is something more.
In the fragment-asperity model, earthquakes result from interactions between fault fragments. The key assumption is the relationship between stress-bearing interaction energy and contact surface area. A linear relationship — energy proportional to surface — is the simplest possibility. The authors prove it is also unique: only linear scaling produces a closed-form expression for total Tsallis entropy as a function of q.
This turns q from a fitting parameter into a diagnostic. The range 1.4 ≤ q ≤ 1.8 corresponds to the region of steepest entropy variation — the system's sensitivity is highest there. That this range coincides with empirical q values from mainshocks worldwide is now not a coincidence of curve fitting but a consequence of the fragment-asperity mechanics being near criticality.
The shift matters. A fitting parameter tells you what the data looks like. A criticality indicator tells you what the system is doing. If q measures proximity to a critical point, then changes in q over time or space would signal changes in the fault system's mechanical state — not just a different statistical description of the same thing.
The uniqueness result is what gives the interpretation force. If multiple functional forms for the energy-surface relationship produced closed-form entropies, q could be an artifact of model choice. Because only one does, the parameter is pinned to the mechanics.
Note (2026-11-11): checked against the full text of the paper (Sotolongo-Costa, Mora-Rodriguez and Mora-Ramos, arXiv 2603.14759, 'Thermodynamic criticality in seismicity: uniqueness of surface-energy scaling in the fragment-asperity model'). The proposal of a linear relation between interaction energy and contact surface, the statement that it is the only relation leading to a closed elementary form for the total entropy as a function of q (with the proof in an appendix), the range 1.4 to 1.8 of strong entropy variation, and its overlap with most reported mainshock values are as in the paper. Two steps in the essay are not. (1) The paper's uniqueness is a mathematical one: only the linear relation gives a closed elementary expression. The essay turns this into 'because only one does, the parameter is pinned to the mechanics' and says that the overlap 'is now not a coincidence of curve fitting but a consequence of the fragment-asperity mechanics being near criticality'. The paper says the coincidence of the range with reported values is 'indicating the role of q as a criticality indicator, more than just a fitting parameter'; a closed form being available does not by itself show the physical relation is linear, and the paper treats the linear relation as a proposal. (2) The paper compares the interval with q values reported in earlier studies of different catalogs; the essay's suggestion that changes in q over time or space would signal the fault's mechanical state is an extrapolation the paper supports only by citing other reports of q rising before large earthquakes.